Alison Redford and the Progressive Conservatives defied the polls and beat Danielle Smith's Wildrose by 10 points Monday night – after trailing by about that much for most of the campaign. But while it may appear the polls bungled the Alberta election, there is some indication a change of heart took place in the last 48 hours before the vote.
It would not have been the first time this year that Albertans swung so dramatically. In polls conducted prior to the official launch of the campaign, the PCs held a two-point lead over Wildrose. But by the end of that first week, Wildrose was up by 11 percentage points.
In the final week of the campaign, the polling was in remarkable agreement. In six polls conducted between Apr. 16 and 21, Wildrose scored between 40 and 41 per cent in every single survey. The Tories were pegged at between 31 and 34 per cent support.
It is very unlikely all of these pollsters could have been coming to the same erroneous conclusions. This was not a matter of a few points, as in the 2011 federal election, that could be explained by a systemic problem in methodology. The odds that all of these polling firms, using different methods and polling over such a short period of time, were missing the mark by as many as 10 points are extremely low.
They also all agreed that Wildrose's support was beginning to sag and that the Tories were experiencing a small upswing. But at the time, it was not enough to point to any Wildrose collapse.
However, a potential smoking gun that points to a last-minute swing in Alberta's voting intentions is a poll conducted by Forum Research on Apr. 22, the day before the vote was held. Forum had been in the field on Apr. 21, along with another polling firm, and both of these surveys had shown the same Wildrose lead that other firms identified through the course of the final week. The Apr. 22 Forum poll, however, showed a swing of seven points between the PCs and Wildrose, with the other parties holding relatively steady.
In short, the gap was 41 to 32 per cent for Wildrose on Saturday, 38 to 36 per cent on Sunday, and finally 44 to 34 per cent on Monday – with the Tories on top. The likelihood the pollsters were generally on the mark in the week before the vote and that Forum captured a last-minute swing (and, as they were the only firm in the field on that last day, they were the only one to be able to record it) seems somewhat high.
A weighted average of all the polls taken up to Apr. 21 shows that Wildrose had the support of 41 per cent of Albertans two days before the vote. The Tories trailed with 32 per cent, the NDP with 12 per cent, and the Liberals with 11 per cent.
Over that last two days, if this swing truly did take place, the Tories picked up 12 points. The majority (seven) came from Wildrose, while the NDP dropped two points and the Liberals one. This suggests that, while some progressive voters did swing to the PCs, the Tory win resulted from Albertans who decided, in the final 48 hours of the campaign, not to cast their ballot for Wildrose after all.
This 12-point swing took place in every part of the province: to 45 from 33 per cent in Calgary, to 44 from 32 per cent in Edmonton, and to 43 from 31 per cent in rural Alberta. The PCs simply gained across the board.
But where these votes came from was not uniform. The largest swing from Wildrose to the Tories took place in Edmonton, where Ms. Smith's party dropped eight points (the Liberals and NDP dropped two apiece). But the Progressive Conservatives were already expected to do very well in the provincial capital. Ms. Redford won the election elsewhere.
In Calgary, the PCs took equally from Wildrose and the NDP. The right-wing party fell five points to 37 from 42 per cent in the metropolitan area, with the NDP dropping to five from 10 per cent. The survival of three Liberal incumbents in the city can be chalked up to their vote holding absolutely firm at 11 per cent over the final days.
In rural Alberta, the Tories stole from everyone: four points each from Wildrose and the NDP and three from the Liberals. But this was the region of the province where Wildrose's support held most strongly, and it is where the vast majority of its MLAs were elected.
There is the possibility the polls somewhat over-estimated Wildrose support in the final week of the campaign and that the swing was not as dramatic as the numbers would suggest. But it seems very likely that Danielle Smith would have won an election held last week – and that a large enough number of Albertans changed their minds and opted for the Tories to swing the election at the very last moment.
Éric Grenier writes about politics and polls at ThreeHundredEight.com.